Most bug reports arise from a Python tracebacks.  Such bugs can be
fixed in minutes.

The present critical bugs are far different: they are "data time
bombs".  The cause of a bug happens far from its discovery.

In such cases, one needs a plan to uncover the actual bug. I developed
such a plan for the unicode bug yesterday on my walk.

The initial plan was simple: instrument every change made in the mass
update.  We want to discover where our expectations do not meet
reality.

This would work, but there is an easier way: instrument every call to
g.u, g.toUnicode, g.cleanSaxInputString, etc. v.bodyString already has
such a check, so maybe I shouldn't have been so disgusted with myself
yesterday :-)  But I may have disabled another such check during the
mass update, and that would indeed have been a disgusting mistake.

Anyway, this plan is very likely to uncover the bug.  Terry has
apparently already shown one way to corrupt a file.  We want to put in
place a monitoring scheme that will catch all such bugs.

Edward
-- 
You received this message because you are subscribed to the Google Groups 
"leo-editor" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/leo-editor?hl=en.


Reply via email to